- fix: parse fediverse addresses as text, so they are not mistaken for email addresses (issue #82)
- fix: restrict elements that can appear inside a label for a labeled link (issue #59)
- fix: Generic schemes were linkified even without content & also a lone http:// was linkified (issue #71)
- Add new methods for working with emojis (they are standalone helper functions and not part of message parsing):
parser::is_emoji::emoji
(rust only) - nom parser that eats one emojiparser::is_emoji::get_first_emoji(text)
- get first emoji if text begins with an emojiparser::is_emoji::count_emojis_if_only_contains_emoji(text)
- counts emojis in texts that contain only emojis
- Parse IRI links (Links that contain non ASCII characters in location part) - link parsing is now RFC3987 and RFC3988 compliant.
- upgrade rust toolchain to 1.77.2
- improved example page (added example text)
- fix absolute unix paths being detected as bot commands suggestions
- fix parenthesis in target of labeled link
- fix bot command suggestion with
@
- char was detected as email address
- upgraded nom to 7
- The following generic schemes (schemes that don't end in
://
) get linkified now:mailto:
,news:
,feed:
,tel:
,sms:
,geo:
,maps:
,bitcoin:
,bitcoincash:
,eth:
,ethereum:
,magnet:
- added
scheme
property toLinkDestination
- hashtag parsing per UAX31
- Fixed problem of IPv6 links being detected as punycode
- fixed: Fixed the bug of brackets being parsed as part of the email address(#34)
- support for
<[email protected]>
- fix: code block with emojis removed some chars at the end
- update rust toolchain to
1.60.0
- enable more clippy lints to prevent panics
- fix: do not parse last dot for email addresses #19
()
,[]
,{}
,<>
brackets are now parsed as part of links (but only if they are opened and then closed again properly)
- count brackets in links to fix #12
- fix links eating trailing
.
,,
,;
or:
#13
This version marks the beginning of using the message parser inside of deltacht-desktop.